Recent animal models suggest that enteral feeding (TEN) compared to parenteral nutrition (TPN) improves resistance to infection. This prospective clinical trial examined the impact of early TEN vs. TPN in the critically injured. Seventy-five patients with an abdominal trauma index (ATI) greater than 15 and less than 40 were randomized at initial laparotomy to receive either TEN (Vivonex TEN) or TPN (Freamine HBC 6.9% and Trophamine 6%); both regimens contained 2.5% fat, 33% branched chain amino acids, and had a calorie to nitrogen ratio of 150:1. TEN was delivered via a needle catheter jejunostomy. Nutritional support was initiated within 12 hours postoperatively in both groups, and infused at a rate sufficient to render the patients in positive nitrogen balance. The study groups (TEN = 29 vs TPN = 30) were comparable in age, injury severity and initial metabolic stress. Jejunal feeding was tolerated unconditionally in 25 (86%) of the TEN group. Nitrogen balance remained equivalent throughout the study period, at day 5 TEN = -0.3 +/- 1.0 vs. TPN 0.1 +/- 0.8 gm/day. Traditional nutritional protein markers (albumin, transferrin, and retinol binding protein) were restored better in the TEN group. Infections developed in 5 (17%) of the TEN patients compared to 11 (37%) of the TPN group. The incidence of major septic morbidity was 3% (1 = abdominal abscess) in the TEN group contrasted to 20% (2 = abdominal abscess, 6 = pneumonia) with TPN. This clinical study demonstrates that TEN is well tolerated in the severely injured, and that early feeding via the gut reduces septic complications in the stressed patient.